how to apply function with two matrices and one vector as input variables

Question

I have 2 matrices and 1 vector with which I intend to substitute them into a function fun accepting 3 variables. But I don't know how to get my goal.

The 2 matrices both have 5 columns and the single vector has 5 elements. I intent to run the function row by row and the Nth elements of each row and the Nth element of the vector could be used. the first matrix eta is:

    eta.1    eta.2    eta.3    eta.4    eta.5
 1.103990 1.345141 1.671156 2.041531 2.435917
 1.031078 1.109853 1.229888 1.380356 1.552349
 1.018405 1.065593 1.139852 1.236245 1.349988
 1.013136 1.046822 1.100638 1.171764 1.257230
 1.010249 1.036439 1.078646 1.135048 1.203625
 1.008425 1.029847 1.064566 1.111308 1.168612
 1.007169 1.025289 1.054776 1.094688 1.143918

the second matrix fn is:

     f.1      f.2       f.3       f.4       f.5
 6.27530 15.29211  28.49757  46.41790  69.23123
10.96466 23.60472  39.23650  58.71576  82.53972
14.17965 29.67335  47.61181  68.85091  93.98208
16.78984 34.69621  54.71981  77.67461 104.17505
19.04558 39.07866  61.00511  85.59340 113.45587
21.06106 43.01689  66.70069  92.83912 122.03283
22.89981 46.62361  71.94674  99.55885 130.04526

the vector n is

1 2 3 4 5

As the function runs, the input element are used like:

fun(1.103990, 6.27530,1)
fun(1.345141, 15.29211,2)
...
fun(1.031078,10.96466,1)
fun(1.109853,23.60472,2)

I hope apply fun to these 3 lists in one step in some form like:

xxx( fun, eta, fn, n)

Does this work: matrix(unlist(Map(fun, t(eta), t(fn), n)), nrow = nrow(eta), byrow = TRUE)? — mt1022
What do you need to happen to the three inputs??? What result to do expect from fun(1.103990, 6.27530,1) for example?? What is your main problem? — Onyambu
Is fun vectorized? If so, sapply(seq_along(n), function(i) fun(eta[i, ], fn[i, ], n)). — Rui Barradas

Ameya Ameya · Accepted Answer · 2018-07-15T09:43:47

?mapply may be helpful. Does this work for you?

Below fun is a function that accepts three arguments.

x1 <- read.table(
  text = ' 1.103990 1.345141 1.671156 2.041531 2.435917
  1.031078 1.109853 1.229888 1.380356 1.552349
  1.018405 1.065593 1.139852 1.236245 1.349988
  1.013136 1.046822 1.100638 1.171764 1.257230
  1.010249 1.036439 1.078646 1.135048 1.203625
  1.008425 1.029847 1.064566 1.111308 1.168612
  1.007169 1.025289 1.054776 1.094688 1.143918'
)
x2 <- read.table(
  text = ' 6.27530 15.29211  28.49757  46.41790  69.23123
10.96466 23.60472  39.23650  58.71576  82.53972
14.17965 29.67335  47.61181  68.85091  93.98208
16.78984 34.69621  54.71981  77.67461 104.17505
19.04558 39.07866  61.00511  85.59340 113.45587
21.06106 43.01689  66.70069  92.83912 122.03283
22.89981 46.62361  71.94674  99.55885 130.04526'
)
x1 <- t(x1) # Expensive operation in case of large data.frames
x2 <- t(x2) # Expensive operation in case of large data.frames
x3 <- 1:5
fun <- function(x1, x2, x3) {
  sqrt(x1**2 + x2**2 + x3**2)
}
mapply(fun, x1, x2, x3)
[1]   6.449665  15.480892  28.703732  46.634636  69.454279  11.058340  23.715282  39.370237  58.868038  82.705593  14.251302  29.759758
[13]  47.719846  68.978084  94.124672  16.850079  34.769568  54.813037  77.786362 104.302549  19.098553  39.143529  61.088353  85.694332
[25] 113.572370  21.108888  43.075671  66.776608  92.931896 122.140809  22.943751  46.677749  72.016984  99.645185 130.146372

how to apply function with two matrices and one vector as input variables

1 Answers